New methods for drive cycles

Researchers at Linköping University have developed new methods of generating and  transforming drive cycles. “A powerful and effective tool for vehicle  development” says Peter Nyberg, who presents the methods in his doctoral thesis.

Fordonslaboratoriet, LiUFoto: Peter NybergA drive cycle simulates a vehicle's speed at different times over a given route and is used by car manufacturers to measure fuel consumption and emission levels and optimize the control of the powertrain. The test values are ​​used by vendors to state the fuel consumption, as well as for calculating the vehicle tax.

Real-world conditions

Peter Nyberg is a doctoral student in Vehicular Systems at Linköping University. His doctoral thesis presents methods for generating representative drive cycles - cycles that reflect real-world conditions - and transforming existing cycles to meet specific conditions. The methods can be combined to get test results that are easier to compare.

“The drive cycles used today are not always up to the mark”, says Peter Nyberg, “and for future vehicles, they will be outdated. The methods we have developed form a powerful and effective tool for vehicle development.”

In order to optimize the control of the powertrain in terms of robustness and sensitivity it is necessary to run several equivalent drive cycles on the same vehicle for comparison. Designing drive cycles with special conditions is a new step in vehicle technology and means that such comparisons can be made faster and yield more reliable results. Peter Nyberg explains:
Peter Nyberg
“Previously, a drive cycle had to be scaled up in terms of time or speed, or repeated several times to make it more demanding for the vehicle to follow. With our methods, you can start with an existing cycle in a database and create new ones which have the same characteristics in the positions you want to study. If a number of such different but equivalent cycles give different test results, for example in terms of fuel consumption, we can make a much more efficient and reliable analysis of the causes than has been possible before.”

Mathematical model

To generate drive cycles, the researchers have used a mathematical model, called a Markov chain, whose parameters are estimated from real driving. The transformation of the drive cycles have been set up as an optimization problem where various requirements for driving cycle characteristics can be formulated.

The possibility to change details of a drive cycle opens for more elaborate studies of different vehicle types, such as studies of hybrid electric vehicles’ dependence on braking energy or customization of drive cycles for concept cars that cannot follow standard drive cycles. As our driving behavior changes due to changes in the infrastructure and the vehicle fleet, the automotive industry will see an increasing need for new drive cycles.

Peter Nyberg presents his thesis "Evaluation, Generation and Transformation of Driving Cycles" on June 10 at Linköping University.

The project "Drive cycles" is part of the Swedish Electric and Hybrid Vehicle Center and is conducted as two PhD projects, one at Linköping University and the other at Chalmers University of Technology.

Text: Emilia Lundgren, Swedish Electric and Hybrid Vehicle Center

The drive cycles can be followed in the chassis dynamometer at the Vehicle laboratory, Linköping University. Picture: Peter Nyberg